Intelligent Sampling Strategy for Online Monitoring of Complex Systems

Industrial Problems Seminar

Honghan Ye
Solventum

Abstract

The rapid progress in sensor technology has led to a data-rich environment, offering significant potential for enhancing detection capabilities in complex systems. However, the vast amount of data, consisting of heterogeneous, high-dimensional streams with high sampling frequencies, results in substantial costs for data collection, transmission, and analysis. As a result, resource limitations often restrict data observability to only a subset of these streams at any given time, presenting considerable challenges in many real-time monitoring applications. In this talk, I will explore recent advancements in intelligent sampling strategies designed to improve online process monitoring and quality enhancement.